With the increased popularity of interactive camera-based computing systems, the number of camera-based programs being written has increased dramatically. However, developing such programs with conventional integrated development environments (IDEs) poses unique challenges. For example, in many conventional IDEs the programmer is presented with discrete textual data values rather than continuously visual representations that more accurately reflect the interactive computer vision data relied on in camera-based computing. Moreover, replicating inputs for debugging in a frame-based loop architecture typically associated with camera-based programming is difficult.